Unverified70% confidenceFactExact time
The teacher model in distillation does not have to be larger than the student model; it can be a domain expert model of comparable size.
2
Sources
70%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
Verified知识蒸馏由Hinton等人在2015年提出，核心思想是让小型学生模型模仿大型教师模型的输出分布76% similarUnverifiedKnowledge distillation, proposed by Hinton et al., has student models inherit implicit knowledge by learning from the soft probability distributions output by teacher models.63% similarUnverifiedKnowledge distillation was first proposed by Hinton, Vinyals, and Dean at Google in 2015, with the core idea of having a small student model learn from the soft labels output by a large teacher model.63% similarUnverified卡内基梅隆大学开发的Cognitive Tutor引入产生式规则模型和学生模型，在代数教学中取得约0.4 SD的持续效果，被美国数千所学校部署63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/911344API
curl https://kongchang.com/api/v1/knowledge/claims/911344MCP
get_claim(id=911344)